|
Lecturer(s)
|
-
Matějů Lukáš, Ing. Ph.D.
-
Červa Petr, doc. Ing. Ph.D.
|
|
Course content
|
Lectures: 1. Introduction to machine learning, definition of basic tasks 2. The task of regression, analytical solution, types of regression, regularization. 3. Stochastic gradient descent method, numerical solution of the regression problem. 4. Basics of classification, distance-based methods. 5. Binary linear classification and logistic regression. 6. Method likelihood estimation and learning of the logistic regression model. 7. Multiclass linear classification, softmax. 8. Linear classification using SVM (support vector machine). 9. Nonlinear classification and (deep) neural networks. 10. Neural networks learning, back propagation. 11. Generative classifiers, Bayesian classification. 12. Methods for transformation of feature vectors (LDA, PCA). 13. Decision trees and random forest. 14. Unsupervised learning, clustering (K-means , LBG, SOM).
|
|
Learning activities and teaching methods
|
Monological explanation (lecture, presentation,briefing)
- Home preparation for classes
- 60 hours per semester
- Preparation for exam
- 50 hours per semester
- Class attendance
- 40 hours per semester
|
|
Learning outcomes
|
The course is an introduction to machine learning with emphasis on task of classification of objects into classes. Students will learn the basic principles of methods, such as deep neural networks, that are currently used in state-of-the-art systems for processing and/or recognition of speech, image or text.
The course provides a basic theoretical foundation for other courses in the master study, which are focused on processing/recogniton of speech, image or text. Students will learn the basic methods for regression, extraction and transformation of features, classification of objects into classes and cluster analysis.
|
|
Prerequisites
|
Knowledge of mathematics at the bachelor level, programming in Matlab.
|
|
Assessment methods and criteria
|
Combined examination, Written exam
Requirements for obtaing the credit are activities at the practicals /seminars.
|
|
Recommended literature
|
-
CHRISTOPHER BISHOP. Pattern Recognition and Machine Learning. Springer, 2006.
-
Kotek Z., Mařík V., Hlaváč V., Psutka J., Zdráhal Z. Metody rozpoznávání a jejich aplikace. Academia, Praha, 1993..
-
Mařík V., Štěpánková O., Lažanský J. a kol. Umělá inteligence (1), Academia, Praha, 1993..
-
RICHARD DUDA, PETER HART AND DAVID STORK. Pattern Classification. 2nd ed. John Wiley & Sons, 2001.
-
Zdráhal Z., Mařík V. Základní metody prohledávání stavového prostoru. In: Metody umělé inteligence a expertní systémy II, ČSVTS FEL ČVUT, Praha, 1985, s.1-35..
|